Double Minute Chromosomes Harboring MDM2 Amplification in A

Double minute chromosomes harboring MDM2 amplification in a pediatric atypical lipomatous tumor Bérengère Dadone-Montaudié, Fanny Burel-Vandenbos, Christine Soler, Olivier Rosello, Corinne Boyer, Thibault Fabas, Laurence Bianchini, Florence Pedeutour

To cite this version:

Bérengère Dadone-Montaudié, Fanny Burel-Vandenbos, Christine Soler, Olivier Rosello, Corinne Boyer, et al.. Double minute chromosomes harboring MDM2 amplification in a pediatric atypical lipomatous tumor. Genes, Chromosomes and Cancer, Wiley, 2019. hal-02398624

HAL Id: hal-02398624 https://hal-cnrs.archives-ouvertes.fr/hal-02398624 Submitted on 7 Dec 2019

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Genes, Chromosomes & Cancer

Double minute chromosomes harboring MDM2 amplification in a pediatric atypical lipomatous tumor

Journal: Genes, Chromosomes and Cancer

Manuscript ID GCC-18-0279.R2

Wiley - Manuscript type: Brief Report

Atypical lipomatous tumor, pediatric, double minute chromosome, well Keywords: differentiated liposarcoma, dedifferentiated liposarcoma For Peer Review

John Wiley & Sons Page 1 of 18 Genes, Chromosomes & Cancer

1 2 3 Double minute chromosomes harboring MDM2 amplification 4 5 6 in a pediatric atypical lipomatous tumor 7 8 9 10 Bérengère Dadone-Montaudié 1,2, Fanny Burel-Vandenbos 2, Christine Soler 3, Olivier 11 12 Rosello 4, Corinne Boyer 5, Thibault Fabas 1, Laurence Bianchini 1, Florence Pedeutour 1 13 14 15 16 1 Laboratory of Solid Tumors Genetics, Institute for Research on Cancer and Aging of Nice 17 18 (IRCAN) CNRS UMR 7284/INSERM U1081, Université Côte d’Azur (UCA), Centre 19 20 Hospitalier Universitaire de Nice, Faculté de Médecine, 28 avenue de Valombrose 06000 Nice, 21 For Peer Review 22 23 France 24 25 26 2 Central Laboratory of Pathology, Nice University Hospital, Hôpital Pasteur, 30 avenue de la 27 28 29 Voie Romaine 06000 Nice, France 30 31 3 Department of Pediatric Onco-hematology, Nice University Hospital, Hôpital Archet 2, 151 32 33 route de Saint-Antoine, 06200 Nice, France 34 35 4 Department of Pediatric Surgery, Hôpital Lenval, 57 Avenue de la Californie, 06200 Nice, 36 37 38 France 39 40 5 Department of Pediatric Radiology, Hôpital Lenval, 57 Avenue de la Californie, 06200 Nice, 41 42 France 43 44 45 Corresponding author: Florence Pedeutour, Laboratoire de Génétique des Tumeurs Solides, 46 47 Faculté de Médecine, 28 avenue de Valombrose, 06000 Nice. Phone number: +33493377012. 48 49 Fax number: +33492037529. 50 51 52 Email: [email protected] 53 54 55 56 Funding: This work was supported by the Institut National du Cancer (INCa) and the Direction 57 58 Générale de l’Offre de Soins (DGOS) (PRT-K 2016, FILIPO), the Direction de la Recherche 59 60

1 John Wiley & Sons Genes, Chromosomes & Cancer Page 2 of 18

1 2 3 et de l’Innovation (DRCI) du Centre Hospitalier Universitaire de Nice (prime d’intéressement 4 5 6 à la recherche), the Fondation ARC pour la recherche sur le cancer, the Alan B. Slifka 7 8 Foundation, the GEMLUC (Groupement des Entreprises Monégasques dans la Lutte contre le 9 10 Cancer) and Infosarcomes. BDM was a Fondation ARC pour la recherche sur le cancer 11 12 fellowship recipient at the time of the study. 13 14 15 16 17 18 19 Abstract 20 21 For Peer Review 22 Adipocytic tumors are rare in children and are mostly benign. Less than 25 cases of 23 24 pediatric well-differentiated liposarcoma (WDLPS), atypical lipomatous tumors (ALT) and 25 26 dedifferentiated liposarcoma (DDLPS) have been reported. Among them, only three cases were 27 28 29 genetically analyzed. We describe the genetic features of a rapidly growing adipose tumor that 30 31 occurred in the thigh of a 7-year-old girl. Histologically, it was composed of mature adipocytic 32 33 cells with a few atypia. Molecular analysis showed high-level amplification of the 12q13-21 34 35 region including MDM2 among 64 amplified genes. MDM2 amplification is a diagnostic 36 37 38 hallmark of ALT/WDLPS/DDLPS. In adult cases it is typically located in ring or giant marker 39 40 chromosomes. In the present case, extra-copies of MDM2 were located on double minute 41 42 chromosomes (dmin). This raised the hypothesis of dmin being precursors of adult’s rings and 43 44 45 giant markers and may provide indications for a better understanding of the mechanisms of 46 47 adipose tumor oncogenesis. 48 49 50 51 52 Keywords 53 54 Atypical lipomatous tumor; well-differentiated liposarcoma; dedifferentiated liposarcoma; 55 56 pediatric; double minute chromosome 57 58 59 60

2 John Wiley & Sons Page 3 of 18 Genes, Chromosomes & Cancer

1 2 3 1. INTRODUCTION 4 5 6 The epidemiologic features of adipose tumors show differences between children and 7 8 adults. Adipocytic tumors are very uncommon in children.1 They are mainly represented by 9 10 two benign entities: lipoblastoma, which is almost exclusively a pediatric tumor, and lipoma, 11 12 one of the most frequent tumors of adulthood. Malignant pediatric adipose tumors are even 13 14 1 15 rarer, accounting for 2% of all soft tissue sarcomas in patients younger than 20 years old. 16 17 Myxoid liposarcoma is the most frequent pediatric liposarcoma, followed by pleomorphic 18 19 liposarcoma. Well-differentiated liposarcoma (WDLPS) / atypical lipomatous tumors (ALT) / 20 21 For Peer Review 22 dedifferentiated liposarcoma (DDLPS) appear to be exceptional in children whereas they are 23 24 the most frequent types of adult sarcomas. They usually occur in middle-aged adults, with a 25 26 peak incidence in the sixth decade.1,2 To the best of our knowledge, less than 25 pediatric 27 28 3-11 29 WDLPS/ALT/DDLPS cases have been reported so far. Adult WDLPS/ALT/DDLPS have 30 31 been well characterized at the clinical and biological levels. Genetically, they share 32 33 amplification of the MDM2 gene as a common recurrent and pathognomonic feature. In adult 34 35 WDLPS/ALT/DDLPS, MDM2 amplification is typically located in supernumerary ring or giant 36 37 38 rod marker chromosomes, together with several amplified genes from the 12q13-15 region. 39 40 Additional genes originating from other chromosomal segments (inter-patient variability) are 41 42 often co-amplified; as such they are part of the composition of the large supernumerary 43 44 2 45 chromosomes. Whereas the genetic landscape of adult ALT/WDLPS/DDLPS is now well 46 47 defined, the characteristics of its pediatric counterpart remain almost unexplored. Indeed, the 48 49 status of MDM2 has been investigated in only three of the few reported pediatric cases.3-5 50 51 52 Here we report the first complete chromosomal and genetic description of a pediatric 53 54 ALT that showed both similarities and differences in comparison to adult cases. 55 56 57 58 2. MATERIALS AND METHODS 59 60

3 John Wiley & Sons Genes, Chromosomes & Cancer Page 4 of 18

1 2 3 2.1 Case report 4 5 6 A 7-year-old-girl presented with a rapidly growing, painless mass of the right thigh. 7 8 Magnetic resonance imagering revealed a well-defined, heterogeneous, lipomatous tumor 9 10 measuring 7x7x5cm located in adductor muscle (Fig. 1A). A biopsy was first performed. Two 11 12 months later, following collegial discussions on potential functional consequences of a wide 13 14 15 surgical procedure versus oncologic risks of insufficient tumor removal, it was followed by a 16 17 complete surgical resection (Fig. 1B). No recurrence or metastasis was detected after six months 18 19 of follow-up. Informed consent for molecular analyses from both parents and child has been 20 21 For Peer Review 22 obtained. 23 24 2.2 Pathological, cytogenetic and immunohistochemical analyses 25 26 Tumor samples from the surgical biopsy and the surgical excision were formalin-fixed 27 28 29 and paraffin-embedded (FFPE) according to standard protocols. The tumor fragment obtained 30 31 from the surgical excision was sampled in 17 blocks. In addition, a fresh fragment was used for 32 33 mechanical and enzymatic disaggregation followed by short-term cell cultures (8 days), 34 35 metaphase cells harvest and spreading according to standard cytogenetic procedures. Antigene 36 37 38 retrieval and immunohistochemistry were done on FFPE tissue sections using the Agilent Dako 39 40 Autostainer Link 48 (Les Ulis, France) and antibodies against MDM2 (1/25, pH6, BioSB, Santa 41 42 Barbara, CA) and HMGA2 (1/100, pH9, BioCheck, San Francisco, CA). 43 44 45 2.3 Fluorescence in situ hybridization (FISH) analysis 46 47 FISH analyses were performed on interphase and metaphase cells from short-term 48 49 culture and on FFPE tissue sections from both the biopsy and the surgical excision, using probes 50 51 52 for MDM2 (green signal) and centromere 12 (red signal) (ZytoLight SPEC MDM2/CEN 12 53 54 Dual Color, ZytoVision, Bremerhaven, Germany) and PLAG1 (BACs RP11-242J1 (3’PLAG1; 55 56 green signal) and RP11-97K15 (5’PLAG1; red signal) (Invitrogen, Carlsbad, CA). 57 58 2.4 Comparative genomic hybridization on arrays (CGH-array) study 59 60

4 John Wiley & Sons Page 5 of 18 Genes, Chromosomes & Cancer

1 2 3 Genomic DNA was extracted from FFPE tissues of the biopsy using a Promega 4 5 6 Maxwell 16 FFPE kit (Promega, Madison, USA). Human reference DNA was provided by 7 8 Agilent Technologies (Santa Clara, CA). Both DNAs were labeled using the Genomic DNA 9 10 SureTag Labeling kit and co-hybridized onto a 4x180K SurePrint G3 Human CGH microarray 11 12 (Agilent Technologies). The microarray was scanned using a SureScan scanner (Agilent 13 14 15 Technologies). Data were analyzed using Agilent Cytogenomics software (version 2.9.2.4, 16 17 Agilent Technologies) and expressed according to the hg19 reference genome (GRCh37, 18 19 Genome Reference Consortium Human Build 37). 20 21 For Peer Review 22 2.5 Single-nucleotide polymorphism on arrays (SNP-array) study 23 24 Genomic DNA was extracted from FFPE tissues of the surgical sample. Copy number 25 26 alterations and loss of heterozygosity were evaluated using Affymetrix OncoScan CNV FFPE 27 28 29 Assay with 80ng input of genomic DNA (Affymetrix, Santa Clara, CA). Experimental 30 31 procedures were performed according to the manufacturer’s recommendations (Affymetrix). 32 33 Data were analyzed using the Chromosome Analysis Suite (ChAS 3.3) software (ThermoFisher 34 35 Scientific, Waltham, MA). Annotations were based on the human reference hg19 (Genome 36 37 38 Reference Consortium Human Build 37 (GRCh37)). 39 40 2.6 Targeted next generation sequencing (targeted-NGS) analysis 41 42 Targeted sequencing was performed using the Ion Torrent semiconductor-based 43 44 45 sequencing technology (Life Technologies, Grand Island, NY). Libraries from the biopsy 46 47 specimen were barcoded with a sample-specific 10nt-barcode sequence (Ion Xpress barcode 48 49 adapter kit) and processed on a 318 v2 chip. Emulsion PCR (Ion One Touch 2, Thermo Fisher 50 51 52 Scientific, Waltham, MA) was performed according to the manufacturer’s recommendations. 53 54 Sequencing was performed using the Ion Torrent Personal Genome Machine by targeting 50 55 56 genes with the Ion AmpliSeq Cancer Hotspots Panel v2. The data were analyzed using Torrent 57 58 Suite software (version 5.6) and the Torrent Variant Caller plugin with the “somatic-low 59 60

5 John Wiley & Sons Genes, Chromosomes & Cancer Page 6 of 18

1 2 3 stringency” parameters. Variants were annotated using Ion Reporter software (version 5.0) and 4 5 6 then expressed according to the hg19 reference genome (GRCh37, Genome Reference 7 8 Consortium Human Build 37). 9 10 11 12 3. RESULTS 13 14 15 3.1 Pathological findings 16 17 Analysis of the biopsy specimen revealed a proliferation of mature adipocytes. No 18 19 evidence of nuclear atypia was detected. Macroscopically, the complete surgical resection 20 21 For Peer Review 22 tumor specimen consisted of a yellowish, well-circumscribed and encapsulated monobloc 23 24 measuring 10x8x7cm (Fig. 1B). No necrotic or hemorrhagic area was detected. 25 26 Microscopically, a few large and hyperchromatic nuclei were observed within a proliferation 27 28 29 of mature adipocytes mixed with thick fibrous septa (Fig. 1C-E). Immunostaining showed a 30 31 strong nuclear expression of MDM2 (Fig. 1F) and HMGA2 (Fig. 1G). 32 33 3.2 Cytogenetic and molecular findings 34 35 MDM2 amplification was first detected in interphase cells from the biopsy sample. More 36 37 38 than 20 extra copies per cell were observed (Fig. 2A). Strikingly, the signals were diffusely 39 40 spread all over the nucleus, in contrast to the clusters usually observed in adult 41 42 ALT/WDLPS/DDLPS (Fig. 2B). Both CGH-array and SNP-array analyses confirmed the 43 44 45 amplification of MDM2 (mean Cy5/Cy3 log ratio: 3.5; average copy number: 22). A large 46 47 discontinuous 12q13-21 amplicon, was observed (from 53,036,089 up to 91,580,752 according 48 49 to nucleotidic coordinates in human reference hg19 GRCh37). Overall, 64 genes were amplified 50 51 52 (Supplementary Fig. 1). Notably, CDK4, FRS2 and CPM and a portion of HMGA2 (5’ region) 53 54 were co-amplified with MDM2. This amplicon was surrounded by deleted regions at 12q11-12 55 56 and 12q22-24.3, respectively (Fig. 2E and Supplementary Fig. 1). The SNP-array results 57 58 indicated an average of 23 copies of the MDM2 gene, and only two values (instead of the three 59 60

6 John Wiley & Sons Page 7 of 18 Genes, Chromosomes & Cancer

1 2 3 expected) of Allele Difference: 2.997 and -2.997 and of B Allele Frequency (BAF): 0.003 and 4 5 6 0.99, respectively (Fig. 2F). This suggested a monoallelic origin of the amplified regions. The 7 8 only other quantitative alteration detected was a loss of 10q11.21-q26.3 (Fig. 2E). FISH 9 10 analysis on interphase cells from the surgical sample confirmed the MDM2 amplification. No 11 12 rearrangement of PLAG1 was detected. Only a few metaphase cells were obtained from the 13 14 15 short-term cell culture. Even though the quality of R-banded cell metaphases was not sufficient 16 17 to establish a full karyotype, it was possible to observe the presence of double minutes 18 19 chromosomes (dmin) (data not shown). No giant marker or supernumerary ring chromosome 20 21 For Peer Review 22 was detected. FISH analysis showed that MDM2 amplification was located on dmin, as 23 24 suspected from features observed on interphase nuclei (Fig. 2C-D). According to the 25 26 International System for Human Cytogenomic Nomenclature (ISCN 2016)12, the molecular 27 28 29 results were: ish dmin(MDM2amp). nuc ish (D12Z3x2,MDM2amp). 30 31 arrGRCh3710q11.21q26.3(45374097_135434303)x1, 12q11q13.3(37902987_57838013)x1, 32 33 12q13.13q21.1(53036089_75664046)x22, 12q21.1q22(75682795_94826994)x1, 34 35 36 12q21.33(91317952_91580752)x21, 12q22q24.33(95364816_133818115)x1. No pathogenic 37 38 mutation was detected using the NGS panel of 50 genes, including TP53 and PTEN. 39 40 41 42 4. DISCUSSION 43 44 45 In adults, the standard treatment of ALT/WDLPS consists in complete surgical resection 46 47 of the tumor.2 In terms of prognosis, anatomic location is a major factor since tumors localized 48 49 in the retroperitoneum are more prone to recurrence and dedifferentiation. Because 50 51 52 ALT/WDLPS presents risks of recurrence, dedifferentiation and metastasis, a careful follow- 53 54 up is mandatory. Considering the rarity of pediatric adipose tumors and the lack of hindsight 55 56 regarding their clinical evolution, patient management can be difficult. The present description 57 58 59 of a novel case of pediatric ALT confirms that malignant well-differentiated adipose tumors do 60

7 John Wiley & Sons Genes, Chromosomes & Cancer Page 8 of 18

1 2 3 occur in children. Checking of MDM2 status, even in the absence of evidence of nuclear atypia, 4 5 6 is therefore highly recommended for any large or deep-seated or rapidly growing pediatric 7 8 adipose tumor. Such MDM2 analysis is crucial to ensure an accurate diagnosis. Notably ALT 9 10 may be misdiagnosed as lipoma in the absence of molecular analysis. According to the current 11 12 World Health Organization classification of soft tissue and bone tumors, MDM2 amplification 13 14 15 is a consistent distinctive feature separating WDLPS/ALT from benign adipocytic tumors, such 16 17 as lipoma and lipoblastoma.2 An appropriate method for this detection is currently FISH 18 19 analysis. Indeed, FISH allows a rapid quantitative determination of MDM2. It provides 20 21 For Peer Review 22 information on the level of MDM2 amplification in a large number of individual cells. Both 23 24 immunohistochemistry and microarrays (CGH-array, SNP-array) may overlook MDM2 25 26 amplification: immunostaining for MDM2 is sometimes unreliable while microarray profiles 27 28 29 can underestimate the level of MDM2 amplification in case of tumor heterogeneity. Indeed, if 30 31 only a small proportion of cells of the tumor sample carry MDM2 amplification, the apparent 32 33 quantitative level of MDM2 may be low. Only three of the 25 pediatric WDLPS/ALT/DDLPS 34 35 reported in the literature have been genetically studied using FISH analysis. 3-5 MDM2 36 37 3 38 amplification was detected in one case of ALT from a 7-year-old girl while low-level 39 40 amplification of CDK4 and no amplification of MDM2 was found in an ALT in a 14-year-old 41 42 girl.4 In the third case, a DDLPS occurring in an 8-year-old girl, Okamoto et al. did not detect 43 44 5 45 MDM2 amplification. Our case and the sole other reported case of MDM2-amplified pediatric 46 47 ALT 3 showed clinical and pathological differences: the case described by Peng et al 3 was a 48 49 predominantly sclerosing deep-seated mass of the face while our case was a well-differentiated 50 51 52 adipocytic tumor located in a limb. In the case described by Peng et al., the nucleic distribution 53 54 of MDM2 signals was not discussed by the authors. In the present case, we performed CGH- 55 56 array and SNP-array in addition to FISH analysis. This allowed confirmation of the high-level 57 58 amplification of MDM2 but also an overview of the quantitative profile of the whole tumor 59 60

8 John Wiley & Sons Page 9 of 18 Genes, Chromosomes & Cancer

1 2 3 genome. A relatively simple genomic profile mainly characterized by 12q13-21 amplification, 4 5 6 similar to those observed in most adult WDLPS/ALT, was detected. In contrast, we noticed 7 8 that the spatial configuration of MDM2 on interphase nuclei of this pediatric tumor was 9 10 different from the usual layout in clusters of MDM2 signals of adult tumors. These typical 11 12 clusters result from the position of 12q amplicons on giant rod or ring chromosomes which are 13 14 2,13 15 the cytogenetic mark of adult WDLPS/ALT/DDLPS. Only a few exceptions have been 16 17 reported. For instance, in an adult case of DDLPS showing neither rings and large markers nor 18 19 dmin, 12q amplification was scattered and inserted into several chromosomes.14 The presence 20 21 For Peer Review 22 of dmin, in addition to the classical ring or marker chromosomes, has been reported only in a 23 24 few cases.15 FISH analysis was not always performed to identify the origin of these dmin. 16-18 25 26 In four cases of adult WDLPS, the dmin were shown to contain sequences from chromosome 27 28 19,20 29 12. In contrast, in five other cases there was no evidence of sequences of chromosome 12 30 31 or MDM2 in the dmin. 17,19,21-23 In the present case, we observed that MDM2 signals were spread 32 33 all over the nuclei. This prompted us to verify the chromosomal localization of the 34 35 supernumerary MDM2 signals on metaphase cells. MDM2 amplification was located on dmin 36 37 38 instead of being grouped within large marker or rings. Dmin are small paired acentric chromatin 39 40 bodies. They are described as an extrachromosomal vector of gene amplification while 41 42 homogeneously stained regions (hsr) represent the intrachromosomal carriers of 43 44 24,25 45 amplification. In WDLPS/ALT/DDLPS, MDM2 amplification is carried by giant rods or 46 47 ring chromosomes. Strikingly, whereas MDM2 owes its name to the initial discovery of 48 49 amplification in mouse dmin 26, the structure of MDM2 amplification in human has been the 50 51 17,21,27-29 52 most deeply explored in giant rods or ring chromosomes of WDLPS/ALT/DDLPS. 53 54 These odd chromosomes represent an original and very peculiar form of hsr since they are not 55 56 inserted in a regular chromosome. They consist in supernumerary genomically complex 57 58 structures harboring a neocentromere.21 MDM2 amplification has been observed in other 59 60

9 John Wiley & Sons Genes, Chromosomes & Cancer Page 10 of 18

1 2 3 tumors than WDLPS/ALT/DDLPS such as intimal sarcoma, osteosarcoma, glioma, breast 4 5 6 cancer or salivary gland tumors. The chromosomal vector of MDM2 amplification has been 7 8 scarcely studied in human tumors other than WDLPS/ALT/DDLPS. Dmin containing MDM2 9 10 have already been reported in salivary glands tumors and neuroblastoma. 30,31 Our observation 11 12 of MDM2 amplification carried by dmin in a pediatric case of adipose tumor raises the issue of 13 14 15 the mechanism of their generation. The episome model of formation of dmin, which is based 16 17 on excision of a gene from its original locus followed by amplification has been extensively 18 19 studied in vitro.24,32 This model is not consistent with MDM2 amplification in 20 21 For Peer Review 22 ALT/WDLPS/DDLPS because MDM2 is usually not deleted on the pair of chromosomes 12 23 24 that remains in cells harboring rings and large markers. In the present case, MDM2 was not 25 26 apparently deleted from chromosomes 12 (Fig. 2D). Mechanisms of recombination without 27 28 33 29 deletion of original locus have also been described to produce dmin. However, these models 30 31 are more compatible with amplification of a small region to select one or a couple of driver 32 33 genes. It does not explain the generation of large amplicons such as those observed in 34 35 ALT/WDLPS/DDLPS. Another model that may account for the formation of dmin in a variety 36 37 34,35 38 of tumors is chromothripsis. According to Garsed et al. chromothripsis may be the initiating 39 40 event in the formation of the ring chromosomes in ALT/WDLPS/DDLPS.27 Large scale 41 42 sequencing of these chromosomes isolated from several WDLPS and DDLPS cell lines using 43 44 45 flow cytometry was in favor of this mechanism rather than others, such as breakage-fusion- 46 47 bridge. Though not being directly the source of amplification, the major “catastrophic” 48 49 rearrangement caused by chromothripsis phenomenon in 12q and in other chromosomes 50 51 52 simultaneously may be the starting point of further copy number changes leading to 53 54 amplification and to neocentromere formation and integration of telomeric sequences. In this 55 56 model, the formation of dmin might directly follow the initial chromosome shattering and 57 58 precede reassembly in rings and then, large marker chromosomes. In the present case, both the 59 60

10 John Wiley & Sons Page 11 of 18 Genes, Chromosomes & Cancer

1 2 3 heterogeneity of dmin (a few of them did not contain MDM2: see Fig. 2D) and the SNP-array 4 5 6 profile indicating an alternation of LOH, monoallelic amplification and preserved regions could 7 8 be in favor of a chromothripsis mechanism of formation of dmin chromosomes.35,36 9 10 Our findings of dmin bearing MDM2 amplification in a pediatric case raise the issue of the 11 12 potential evolution of these dmin if the tumor had not been surgically removed. We cannot 13 14 15 predict whether these dmin would have aggregated leading to ring and large markers similar to 16 17 those observed in most adult cases of WDLPS/ALT/DDLPS. Former observations in 18 19 neuroblastomas suggested that dmin containing MYCN and MDM2 could evolve towards hsr in 20 21 For Peer Review 31 37 22 human tumors. Similar hypotheses have been raised in lung cancer cell lines. However, in 23 24 vivo proof of such an aggregation of dmin resulting in hsr in primary tumors is missing. 25 26 Alternatively, we cannot exclude the possibility that they would have remained as dmin and 27 28 29 constitute a distinct entity from adult cases. Prospective exhaustive genetic studies of these rare 30 31 pediatric lipomatous tumors may help resolve these issues. 32 33 34 35 Acknowledgments: The authors are thankful to Roger Grattery, Annie-Claude Peyron, 36 37 38 Frédérique Keslair and Audrey Bazin for their technical assistance and Jean-François Michiels 39 40 and Zoé Pedeutour-Braccini for their pathological evaluation. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

11 John Wiley & Sons Genes, Chromosomes & Cancer Page 12 of 18

1 2 3 References 4 5 6 1. Dadone B, Refae S, Lemarié-Delaunay C, Bianchini L, Pedeutour F. Molecular 7 cytogenetics of pediatric adipocytic tumors. Cancer Genet. 2015;208(10):469-481. 8 2. Dei Tos AP, Pedeutour F. Atypical lipomatous tumour. In: Fletcher DM, Bridge JA, 9 Hogendoorn PCW, Mertens F, editors. WHO Classification of Tumours of Soft Tissue and 10 Bone. 4th ed. Lyon: IARC; 2013, p. 33–6. 11 3. Peng R, Chen H, Yang X, et al. A novel sclerosing atypical lipomatous tumor/well- 12 differentiated liposarcoma in a 7-year-old girl: report of a case with molecular confirmation. 13 14 Hum Pathol. 2018;71:41-46. 15 4. Kuhnen C, Mentzel T, Fisseler-Eckhoff A, Debiec-Rychter M, Sciot R. Atypical 16 lipomatous tumor in a 14-year-old patient: distinction from lipoblastoma using FISH analysis. 17 Virchows Arch Int J Pathol. 2002;441(3):299-302. 18 5. Okamoto S, Machinami R, Tanizawa T, Matsumoto S, Lee G-H, Ishikawa Y. 19 Dedifferentiated liposarcoma with rhabdomyoblastic differentiation in an 8-year-old girl. 20 Pathol Res Pract. 2010;206(3):191-196. 21 For Peer Review 22 6. Huh WW, Yuen C, Munsell M, et al. Liposarcoma in children and young adults: a multi- 23 institutional experience. Pediatr Blood Cancer. 2011;57(7):1142-1146. 24 7. Alaggio R, Coffin CM, Weiss SW, et al. Liposarcomas in young patients: a study of 82 25 cases occurring in patients younger than 22 years of age. Am J Surg Pathol. 2009;33(5):645- 26 658. 27 8. Debelenko LV, Perez-Atayde AR, Dubois SG, et al. p53+/mdm2- atypical lipomatous 28 29 tumor/well-differentiated liposarcoma in young children: an early expression of Li-Fraumeni 30 syndrome. Pediatr Dev Pathol Off J Soc Pediatr Pathol Paediatr Pathol Soc. 2010;13(3):218- 31 224. 32 9. Stanelle EJ, Christison-Lagay ER, Sidebotham EL, et al. Prognostic factors and survival 33 in pediatric and adolescent liposarcoma. Sarcoma. 2012;2012:870910. 34 10. Hahn HP, Fletcher CDM. Primary mediastinal liposarcoma: clinicopathologic analysis 35 of 24 cases. Am J Surg Pathol. 2007;31(12):1868-1874. 36 37 11. Boland JM, Colby TV, Folpe AL. Liposarcomas of the mediastinum and thorax: a 38 clinicopathologic and molecular cytogenetic study of 24 cases, emphasizing unusual and 39 diverse histologic features. Am J Surg Pathol. 2012;36(9):1395-1403. 40 12. McGowan-Jordan J, Simons A, Schmid M, eds. An International System for Human 41 Cytogenomic Nomenclature. Basel, Switzerland: Karger; 2016. 42 43 13. Dei Tos AP, Marino-Enriquez A, Pedeutour F, Rossi S. Dedifferentiated liposarcoma. 44 In: Fletcher DM, Bridge JA, Hogendoorn PCW, Mertens F, editors. WHO Classification of 45 46 Tumours of Soft Tissue and Bone. 4th ed. Lyon: IARC; :37-38. 47 14. Mandahl N, Magnusson L, Nilsson J, et al. Scattered genomic amplification in 48 dedifferentiated liposarcoma. Mol Cytogenet. 2017;10:25. 49 15. Mitelman Database of Chromosome Aberrations and Gene Fusions in Cancer (2019). 50 Mitelman F, Johansson B and Mertens F (Eds.). 51 http://cgap.nci.nih.gov/Chromosomes/Mitelman. Accessed February 26, 2019. 52 53 16. Bassett MD, Schuetze SM, Disteche C, et al. Deep-seated, well differentiated 54 lipomatous tumors of the chest wall and extremities: the role of cytogenetics in classification 55 and prognostication. Cancer. 2005;103(2):409-416. 56 17. Gisselsson D, Höglund M, Mertens F, et al. The structure and dynamics of ring 57 chromosomes in human neoplastic and non-neoplastic cells. Hum Genet. 1999;104(4):315-325. 58 18. Sreekantaiah C, Karakousis CP, Leong SP, Sandberg AA. Cytogenetic findings in 59 liposarcoma correlate with histopathologic subtypes. Cancer. 1992;69(10):2484-2495. 60

12 John Wiley & Sons Page 13 of 18 Genes, Chromosomes & Cancer

1 2 3 19. Sozzi G, Minoletti F, Miozzo M, et al. Relevance of cytogenetic and fluorescent in situ 4 hybridization analyses in the clinical assessment of soft tissue sarcoma. Hum Pathol. 5 6 1997;28(2):134-142. 7 20. Pilotti S, Della Torre G, Mezzelani A, et al. The expression of MDM2/CDK4 gene 8 product in the differential diagnosis of well differentiated liposarcoma and large deep-seated 9 lipoma. Br J Cancer. 2000;82(7):1271-1275. 10 21. Pedeutour F, Suijkerbuijk RF, Forus A, et al. Complex composition and co- 11 amplification of SAS and MDM2 in ring and giant rod marker chromosomes in well- 12 differentiated liposarcoma. Genes Chromosomes Cancer. 1994;10(2):85-94. 13 14 22. Dal Cin P, Kools P, Sciot R, et al. Cytogenetic and fluorescence in situ hybridization 15 investigation of ring chromosomes characterizing a specific pathologic subgroup of adipose 16 tissue tumors. Cancer Genet Cytogenet. 1993;68(2):85-90. 17 23. Nilsson M, Meza-Zepeda LA, Mertens F, Forus A, Myklebost O, Mandahl N. 18 Amplification of chromosome 1 sequences in lipomatous tumors and other sarcomas. Int J 19 Cancer. 2004;109(3):363-369. 20 24. Storlazzi CT, Lonoce A, Guastadisegni MC, et al. Gene amplification as double minutes 21 For Peer Review 22 or homogeneously staining regions in solid tumors: origin and structure. Genome Res. 23 2010;20(9):1198-1206. 24 25. Turner KM, Deshpande V, Beyter D, et al. Extrachromosomal oncogene amplification 25 drives tumour evolution and genetic heterogeneity. Nature. 2017;543(7643):122-125. 26 26. Fakharzadeh SS, Trusko SP, George DL. Tumorigenic potential associated with 27 enhanced expression of a gene that is amplified in a mouse tumor cell line. EMBO J. 28 29 1991;10(6):1565-1569. 30 27. Garsed DW, Marshall OJ, Corbin VDA, et al. The architecture and evolution of cancer 31 neochromosomes. Cancer Cell. 2014;26(5):653-667. 32 28. Pedeutour F, Forus A, Coindre JM, et al. Structure of the supernumerary ring and giant 33 rod chromosomes in adipose tissue tumors. Genes Chromosomes Cancer. 1999;24(1):30-41. 34 29. Sirvent N, Forus A, Lescaut W, et al. Characterization of centromere alterations in 35 liposarcomas. Genes Chromosomes Cancer. 2000;29(2):117-129. 36 37 30. Rao PH, Murty VV, Louie DC, Chaganti RS. Nonsyntenic amplification of MYC with 38 CDK4 and MDM2 in a malignant mixed tumor of salivary gland. Cancer Genet Cytogenet. 39 1998;105(2):160-163. 40 31. Corvi R, Savelyeva L, Amler L, Handgretinger R, Schwab M. Cytogenetic evolution of 41 MYCN and MDM2 amplification in the neuroblastoma LS tumour and its cell line. Eur J 42 Cancer Oxf Engl 1990. 1995;31A(4):520-523. 43 32. Wahl GM. The importance of circular DNA in mammalian gene amplification. Cancer 44 45 Res. 1989;49(6):1333-1340. 46 33. Stark GR. Regulation and mechanisms of mammalian gene amplification. Adv Cancer 47 Res. 1993;61:87-113. 48 34. Stephens PJ, Greenman CD, Fu B, et al. Massive genomic rearrangement acquired in a 49 single catastrophic event during cancer development. Cell. 2011;144(1):27-40. 50 35. Rode A, Maass KK, Willmund KV, Lichter P, Ernst A. Chromothripsis in cancer cells: 51 52 An update. Int J Cancer. 2016;138(10):2322-2333. 53 36. Zhang C-Z, Leibowitz ML, Pellman D. Chromothripsis and beyond: rapid genome 54 evolution from complex chromosomal rearrangements. Genes Dev. 2013;27(23):2513-2530. 55 37. Fukumoto M, Suzuki A, Inazawa J, et al. Chromosomal location and structure of 56 amplicons in two human cell lines with coamplification of c-myc and Ki-ras oncogenes. Somat 57 Cell Mol Genet. 1993;19(1):21-28. 58 59 60

13 John Wiley & Sons Genes, Chromosomes & Cancer Page 14 of 18

1 2 3 FIGURE LEGENDS 4 5 6 7 8 Figure 1. Radiological and pathological features. A. T1-weighted, axial section of Magnetic 9 10 Resonance Imagery with fat saturation sequences showing an extinction of the hypersignal of 11 12 the adipocytic tumor (arrow). The tumor was deep-seated in the adductor muscle. B. Gross 13 14 15 aspect of the resection: encapsulated and adipose tumor without necrosis or hemorrhage. C. 16 17 Histologically, the tumor was composed of a proliferation of mature adipocytes with thick 18 19 fibrous septa. D. High power view of the fibrous septa showing some striated muscle fibers 20 21 For Peer Review 22 (arrow). E. High power view of the mature adipocytes showing a few atypia with 23 24 hyperchromatic and large nuclei. Immunohistochemical staining showing strong nuclear 25 26 expression of MDM2 (F) and HMGA2 (G). 27 28 29 30 31 Figure 2. Genetics features. A-D. Fluorescence In Situ Hybridization (FISH) analyses using 32 33 probes for MDM2 (green signals) and centromere 12 (red signals). Interphase nuclei and 34 35 metaphase chromosomes are counterstained in blue (DAPI). A. In interphase nuclei, the two 36 37 38 red signals indicated the presence of two centromeres of chromosome 12. The MDM2 39 40 amplification was diffuse, homogeneously distributed all over the nuclei; this distribution 41 42 suggested the presence of double minute (dmin) chromosomes. B. Example of a classical cluster 43 44 45 of MDM2 amplified signals in an adult’s atypical lipomatous tumor. C. Yellow circles focus 46 47 on numerous dmin chromosomes present in a few metaphase cells of the pediatric ALT. D. 48 49 Results of FISH analysis demonstrated that the MDM2 amplification was carried by the dmin 50 51 52 chromosomes; Two chromosome 12 are indicated by white arrows; Orange arrows indicated 53 54 few dmin without MDM2 signals. E. Quantitative whole genome profile representation of 55 56 Single Nucleotide Polymorphism (SNP) array analysis showing high-level large and 57 58 discontinuous amplicon surrounded by a loss of 12q11-12 and 12q22-24.3 regions (blue 59 60

14 John Wiley & Sons Page 15 of 18 Genes, Chromosomes & Cancer

1 2 3 arrowhead) and a loss of the long arm of chromosome 10 (black arrowhead). F. SNP-array 4 5 6 results showing a focus on the 12q region (genomic coordinates from 64.470.744 to 7 8 73.970.757). The figure was centered by the MDM2 gene (black arrowhead). The red bars 9 10 represented losses while the blue bar, gains or amplification. The Allele Difference and the B 11 12 Allele Frequency represented the variation of the heterozygosity. 13 14 15 16 17 Supplementary Figure 1. Genes amplified in the 12q amplicons. We observed discontinuous 18 19 amplicons located on 12q13.13-15 and on 12q21.1-21.33, including several genes and notably 20 21 For Peer Review 22 MDM2, CDK4, HMGA2, CPM, FRS2 and YEATS4. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

15 John Wiley & Sons Genes, Chromosomes & Cancer Page 16 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 26 27 28 29 30 31 Figure 1. Radiological and pathological features. A. T1-weighted, axial section of Magnetic Resonance Imagery with fat saturation sequences showing an extinction of the hypersignal of the adipocytic tumor 32 (arrow). The tumor was deep-seated in the adductor muscle. B. Gross aspect of the resection: encapsulated 33 and adipose tumor without necrosis or hemorrhage. C. Histologically, the tumor was composed of a 34 proliferation of mature adipocytes with thick fibrous septa. D. High power view of the fibrous septa showing 35 some striated muscle fibers (arrow). E. High power view of the mature adipocytes showing a few atypia with 36 hyperchromatic and large nuclei. Immunohistochemical staining showing strong nuclear expression of MDM2 37 (F) and HMGA2 (G). 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Page 17 of 18 Genes, Chromosomes & Cancer

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Figure 2. Genetics features. A-D. Fluorescence In Situ Hybridization (FISH) analyses using probes for MDM2 35 (green signals) and centromere 12 (red signals). Interphase nuclei and metaphase chromosomes are 36 counterstained in blue (DAPI). A. In interphase nuclei, the two red signals indicated the presence of two 37 centromeres of chromosome 12. The MDM2 amplification was diffuse, homogeneously distributed all over 38 the nuclei; this distribution suggested the presence of double minute (dmin) chromosomes. B. Example of a 39 classical cluster of MDM2 amplified signals in an adult’s atypical lipomatous tumor. C. Yellow circles focus on 40 numerous dmin chromosomes present in a few metaphase cells of the pediatric ALT. D. Results of FISH 41 analysis demonstrated that the MDM2 amplification was carried by the dmin chromosomes; Two chromosome 12 are indicated by white arrows; Orange arrows indicated few dmin without MDM2 signals. E. 42 Quantitative whole genome profile representation of Single Nucleotide Polymorphism (SNP) array analysis 43 showing high-level large and discontinuous amplicon surrounded by a loss of 12q11-12 and 12q22-24.3 44 regions (blue arrowhead) and a loss of the long arm of chromosome 10 (black arrowhead). F. SNP-array 45 results showing a focus on the 12q region (genomic coordinates from 64.470.744 to 73.970.757). The figure 46 was centered by the MDM2 gene (black arrowhead). The red bars represented losses while the blue bar, 47 gains or amplification. The Allele Difference and the B Allele Frequency represented the variation of the heterozygosity. 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons Genes, Chromosomes & Cancer Page 18 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 John Wiley & Sons